Microsoft Preps Faster Web Searches
Page 1 of 1
MOUNTAIN VIEW, CALIF. -- One of the benefits of being the world's largest software company is having the resources to do less time-sensitive research. At its silicon valley facility here, Microsoft researchers are free to dream up all manner of new computing ideas free of the pressures of shipping deadlines.
The Microsoft Research campus, spread across five large buildings, was established in 2001, although the company created the unit 10 years earlier to foster long-term research that's not necessarily connected with a specific product release.
More than 700 researchers, many from academia, work to create new methodologies, interfaces and technologies, which have found their way into practically every Microsoft product, from the XBox to Windows XP. Now, two more new technologies are about to leave the lab.
First up is Wild Thing, a search technology so dubbed because it allows for wildcard searches on MSN by using only a few characters. For example, if you find it difficult to spell the last name of California's actor-turned-governor Arnold Schwarzenegger, entering "ar* sc*w" will find him. An entry of "ar* sc*w mo*" will find films from the former Terminator.
Microsoft originally devised Wild Thing for mobile users, so they wouldn't have to enter so many characters on a cell phone keypad, but the company decided to make it available for everyone after seeing how useful it was.
"We wanted to get away from something that forced you to type so many characters," said Bo Thiesson, a researcher at the lab. "But it turned out so useful that even if you have a real keyboard it can save you a lot of typing and make Internet searches a lot faster."
Wild Thing uses the MSN search indexes and rankings, so the more popular the subject matter, the faster you will find it and the less characters needed. Thiesson expects users will eventually get used to the shorthand method of searching. "If you don't get it right, you have to put in more characters, and eventually you adapt to it, so you know how many characters you need to input it because you know how popular your search is," he said.
Thiesson said he hopes to see Wild Thing eventually deployed on all MSN services, such as the main search engine, MSN Mobile and MSN Virtual Earth. He wouldn't give a release schedule but was "optimistic" about it seeing deployment soon.
As a companion technology, Microsoft is developing a technology called project Nocturnal which taps into a user's bookmarks and previously visited pages to share them with friends in the MSN Messenger buddy list.
This serves two purposes. First, Nocturnal learns the interests of each user and allows a search engine to push the most relevant results to the top of the search list. Second, it lets people share their favorite links virally. Instead of the age-old method of pasting a link into an email or an IM message, your links will now be shared with everyone you choose, and it's all done in the background while you use your computer.
This technology is based on the rather obvious observation that buddies are likely to share common interests. What you and your friends recommend is likely to be more relevant to you, so you would like to see those results before they're returned by the search engine and eliminate false hits, explained Lidong Zhou, a researcher for Microsoft.
Nocturnal has two components. The first is found in MSN Messenger Live beta, which has a shared information folder. Drop whatever you want to share in it and that information is shared with friends on your MSN buddies list. You can tell MSN Messenger not to share anything, share with only a few people, or with everyone, said Zhou. The second part is a toolbar for the Web browser that lets you rate pages and capture links in your shared folder, allowing them to be shared with other people.
Currently, Nocturnal is being targeted at MSN Messenger and Internet Explorer. Zhou said that there is no technical difficulty in getting it working on Firefox and AOL Instant Messenger, but he didn't know if those companies would be interested in the technology in the first place. He wasn't even sure when it would find its way into a Microsoft product.
"As researchers we care more about the research questions we want to answer," he said. "We think the combination of social networks with information filtering is a powerful paradigm. We're more interested in how information propagates in a social network, and how you find buddies who share a common interest."